I know that all of you are following the latest Presidential primary polls with bated breath. The Iowa caucuses are a mere seven months away, after all. If Clinton or Giuliani are surging there now, it must mean that they are about to become unstoppable forces on the road to the nomination.
Well, okay, so Presidential primary polls don't mean much at this stage. But polling is a constant feature of the modern political and journalistic landscape, and much of what is written about or believed about polls is garbage. Enter the Mystery Pollster. He began blogging in the aftermath of the 2004 Presidential election, when there was a great deal of confusion about why exit polls in Ohio and elsewhere had been so badly inaccurate; and he provided some of the only informed, intelligent commentary that could be found anywhere about what happened. Since then, he has become much less mysterious, revealing who he actually is (Mark Blumenthal), starting a full-blown website, and getting some other bloggers to help him out. I don't read it that frequently, but I never cease to be amazed when I do. He and his co-bloggers understand how polling works and what polls do and do not mean, and what they post there is always interesting (if often fairly mathematics-heavy). As an example, consider the examination of the degree to which the Clinton impeachment coincided with a collapse in popularity of Monica as a name for female babies.